Researchers from The University of Texas at Austin and Cognizant AI Labs have made impressive progress. They are passionately committed to figuring out the best solutions to our environmental policy challenges. They are committed to fulfilling the global sustainability goals established by the United Nations. To do so, they utilize evolutionary artificial intelligence (AI) technology to optimize local land use policies. This groundbreaking research leverages 175 years of global land use and carbon storage data. The Act believes in and invests deeply in addressing the urgent, existential challenges of climate change and greenhouse gas emissions.
The collaborative research team, including Olivier Francon, Elliot Meyerson, Clemens Schwingshackl, Jakob Bieker, Hugo Cunha, and Babak Hodjat, employed a systematic approach to explore various policy scenarios. They modeled the economic and environmental impacts of different land use policies. I think their goal was to figure out which of those solutions would be most effective at reducing carbon emissions.
Evolutionary AI in Action
To help frame their efforts, the researchers started their investigation by looking at a range of policy situations. They specifically studied the economic and environmental impacts of each of those scenarios. They utilized an iterative process that involved winnowing out less effective strategies while retaining the most promising options across hundreds or thousands of scenarios. This iterative methodology provided the team the space to test approaches and focus on what style worked best.
Two primary tools underpinned the research: a comprehensive set of global land use data accumulated over centuries and a sophisticated model designed to correlate land use practices with carbon fluxes. By taking a dual approach, the researchers were able to build an effective model of prediction. This powerful model connects location, land use, and carbon emissions through time.
The research produced a prescriptive model. This model is intended to provide guidance to decision-makers looking for the most effective land-use strategies to combat climate change. Most striking to me is that the prescriptor calls for decreasing the area of pasture and cultivated crops and increasing the area under range and secondary forests. All of these amendments have the potential to greatly decrease carbon emissions, around 26.02 tons per hectare. All this reduction would happen in tandem with a projected 28.96% land use change.
Interactive Tool for Decision Makers
To make it easier for people to put their principles into practice, the research team turned their research into an interactive tool designed specifically for decision-makers. On this platform, any user can run through countless scenarios. It allows them to test how alternative incentives might steer land use decisions and in turn, help reduce carbon emissions.
Users can select from multiple different alternative solutions along the “Pareto front.” This idea illustrates the historical tension between competing objectives. Plus, manual sliders allow users to change the values of particular variables and see what the possible results could be according to their preferences.
Risto Miikkulainen, a researcher participating in the larger effort, said the challenge lies in optimizing desired outcomes, while minimizing cost.
“There’s always an outcome you want to optimize for, but there’s always a cost.” – Risto Miikkulainen
This finding highlights the precision needed when pursuing sustainable policy changes to avoid causing negative unintended consequences.
Addressing Land Use Emissions
Land use activities—including agriculture and forestry—not only contribute to climate change but make up almost a quarter of all human-induced greenhouse gas emissions. Whatever the reasons, the findings from this research are hugely significant because they show that there are clear pathways to reducing these harmful impacts while still meeting economic needs.
Daniel Young, another principal researcher on the project, touched on the highs as well as the lows when it comes to land-use change.
“You can obviously destroy everything and plant forests, and that would help mitigate climate change.” – Daniel Young
He emphasized the need to ensure these strategies are balanced.
“But we would have destroyed rare habitats and our food supply and cities. So we need to find a balance and be smart about where we make changes.” – Daniel Young
This view eschews previous anthropocentric doctrine in favor of a more complex, holistic understanding of the connection between ecological and human flourishing.
Recognizing Global Impact
The collaborative research project is one of the first applications of the UN-backed Project Resilience initiative. Its contributions are many beyond its theoretical frameworks. Together, they provide highly actionable insights that can help spur concrete, meaningful environmental progress around the globe.
An older variant of this work achieved remarkable acclaim on the artificial intelligence summit NeurIPS. It subsequently went on to win the “Best Pathway to Impact” award in the Climate Change workshop. This accolade highlights the project’s relevance and potential for real-world impact in addressing one of humanity’s most pressing challenges.